Interview with Dr Anant Sudarshan

Dr Anant Sudarshan is the South Asia Director of the Energy Policy Institute at the University of Chicago. Anant's research spans several aspects of energy and environment policy, including the design of environmental regulation, reducing air pollution, climate change, energy efficiency, electricity access, and renewable energy policy. He has vast experience working with electricity utility companies in India.

A study that Anant is currently carrying out forms part of an EEG-funded project on the effect of smart metering on revenue collection, electricity access and supply. Here, we talk to Anant about the benefits of smart meters and the impact they can potentially have, and ask him about his research and findings so far. He also shares his thoughts on an EEG webinar he recently contributed to on domestic electricity meters.


It is thought that smart meters may be able to improve utilities’ revenue collection. What kind of issues do utilities in India typically face when trying to collect payments from their customers?

Utilities in India have consistently struggled to recover revenues from their consumers, notwithstanding the fact that many of them are only being charged subsidised tariffs. One reason is political and policy decisions to not enforce payments, for example to support poor people during economic downturns. For example, states such as Madhya Pradesh have waived arrears following Covid to provide support to people who may have seen reduced incomes. Many political parties have made subsidised or free electricity part of their manifestos. As a broad principle, such decisions may create a social norm encouraging non-payment, independent of their specific motivation.

Given a status-quo of widespread non-payment, utilities are capacity-constrained in the enforcement actions they can take. Given staff and budget limitations, they must selectively choose which people to disconnect. No matter how this discretion is exercised, partial enforcement reduces the expected costs to non-payment. There are also sometimes constraints on the ability of utilities to distribute and collect on bills, including staff shortfalls. Finally, many consumers are able to illegally connect themselves to the grid and it is possible that strict enforcement of disconnections for partial payments may displace people to these illegal connections, effectively reducing the revenue they provide to zero, without changing total demand.


Can you explain the potential impact smart metering can have in terms of improving utilities’ cost recovery, and thereby electricity reliability and access?

There are several potential benefits to smart meters. They improve the quality of data available to the utility, allowing them to quickly identify consumers who do not pay. They are necessary for good energy accounting, thus enabling better measurement of illegal connections. They also reduce the transaction costs involved in enforcement actions such as disconnections. Finally, they open up the possibility of alternative contracts with consumers, including time-of-day tariffs and pre-paid metering. The first of these can lead to more efficient pricing and the second could resolve the challenge of recovering payments ex-post by ensuring that consumption can only occur once money has been deposited by consumers.


As part of an EEG project on smart metering and electricity access, being led by Jameel Poverty Action Lab, South Asia (J-PAL SA), you are carrying out a large-scale neighbourhood-level randomised-control trial in Haryana in India, looking at the political economy and practical challenges of implementing advanced metering infrastructure (AMI) smart meters. Can you tell us more about the study and what it hopes to achieve?

The study seeks to measure the impacts of introducing real-time billing as well as automatic disconnections using smart meters. We use a randomised control design to quantify these impacts, where a treatment group is compared to a control group which is billed as per the status quo.


What are the benefits of AMI smart meters compared to other types of meters?

AMI meters allow for two-way communication between the utility and the consumer, including real-time data on consumption, as well as allowing features such as pre-paid metering, remote disconnections, and time-of-day pricing. This differs from older analogue meters, which must be manually read, thus incurring significant time, staff, and monetary costs. Such meters also do not allow for more sophisticated payment contracts. An alternative to both AMI and analogue meters is the use of automatic meter reading (AMR) meters, which offer only some of the benefits of AMI metering but are nevertheless an improvement over the analogue technology.


Can you tell us about the smart meter installations in Haryana – what has the process been for rolling them out and have they gone well? Have smart meters been well received by consumers, or have utilities faced resistance?

The process of installing meters has gone relatively smoothly in Haryana, although this is not the universal experience for utilities across India. This utility did not face a lot of resistance to changing meters. With that said, smart meters in Haryana are presently only being used to a limited degree even where they are installed. The primary application has been to automate the meter reading and billing system, which may reduce utility costs but need not affect the consumer experience as much.


Can you share any results on the impact of smart meters on electricity consumption, bill payment, electricity theft, utility revenue or quality of services? Have consumers or utilities seen any benefits yet?

Final results from the study are still underway, though it seems fair to say that most of the benefits of new meters are to improve utility operations with a limited impact on consumers.


Do you see smart meters as a solution to improving cost recovery in their own right, or do they need to be accompanied by other measures?

Based on the Haryana experience, smart meters are unlikely to improve cost recovery purely because they are purchased by the utility and installed in the homes of consumers. This is because their benefits also require policy or operational changes – for instance the decision on how stringently to enforce disconnections, or whether to transition consumers to pre-paid contracts. Smart meters make these options available to regulators and utilities but do not imply they will be used. As such, they are best viewed as an enabling technology, but one that must be accompanied by reforming the politics, economics, and operational practices of utilities. That said, in settings where existing metering is old and functioning imperfectly, the act of installing new technology may significantly improve data quality and result in more accurate bills. These factors may change revenue, although it may be hard to predict in advance whether better measurement moves in the direction of higher or lower bills.


What’s next for you in terms of your research on smart meters and electricity access? What do you think is the biggest research gap in this area? You recently took part in EEG’s webinar: What is the impact of using smart/pre-paid domestic electricity meters? What are your key takeaways from the event?

I am working on another smart meters evaluation with the Indian state of Jammu and Kashmir, where we hope to test pre-paid metering as well. This is a contract option that is enabled by smart meters but was not taken up by Haryana and thus evidence on its impacts on payments would be new and valuable. In another project I am also studying how different enforcement techniques, including behavioural nudges, can be targeted and deployed by utilities that are using traditional meters, so as to improve revenue recovery using the tools and hardware currently available. I also have projects considering the impact of increasing hours of supply on enterprise activity and firm profits in the state of Bihar in India. The focus of this paper is on examining impacts on small and informal businesses in rural and semi-urban areas.